In the state of the art, field devices—especially measuring devices—are known, which output signals and especially measured values as 4 . . . 20 mA signals. When an error is present in the field device, a so-called error signal is then output, which usually lies outside of the actual signal range between 4 and 20 mA. The error signal thus lies either below 4 mA or above 20 mA.
In the context of the idea of self-checking field devices, it is, in such case, also required that the device should be able to signal such an error current. In such case, a problem resides, however, in the fact that this error signal should itself not reach the output of the device, since this involves a test and not the presence of such an error. As a simple solution, corresponding error signals are consequently intentionally produced by the field devices in test time periods. However, in these periods of time, a normal process operation is, consequently, not possible.